s1t5

17 ms·

121.
by s1t5 6y ago
The article is just vaguely complaining about undefined problems which makes it very difficult to defend or argue against. Are we talking about ML? Optimization problems? Operational research in general? Automation? And which tasks in which
122.
by s1t5 6y ago
> It actually means that the first player always loses on average if he bets (expected utility -1/2 if not mistaken). If he folds however the expected utility is clearly zero since half of the time wins and half of the time loses. &
123.
by s1t5 6y ago
Watson is just a brand prefix at this point - it goes in front of a number of products like Watson Studio, Watson Machine Learning, Watson Assistant and so on.
124.
by s1t5 6y ago
And it's similar for big cities - local purchasing power in London is lower than New York, San Francisco and LA.
125.
by s1t5 6y ago
The "we are being paid too much" sentiment is mostly a US thing. In Europe the salaries are nowhere nearly as high.
126.
by s1t5 6y ago
> Last but not least. Feel responsible from day one. You are a part of the team and you are responsible for every failure and every success. More or less. If there is a pain in somewhere in your team, feel it in your heart. In a new role
127.
by s1t5 6y ago
Isn't hinting the worst approach in that case though? I can't imagine that the additional uncertainty helped with staff morale.
128.
by s1t5 6y ago
> it seems skeevy techbro at best It really is. But it can still make a lot of sense for very small teams who are highly invested and/or own the company/product outright. If you're putting in these hours as a salaried empl
129.
by s1t5 6y ago
This + willingness to work absurd hours.
130.
by s1t5 6y ago
You can likely land a pretty decent job as it is. People with advanced math degrees who can also program are in demand. Or at least that was the case pre-virus. If you want to take it easier and apply for a job when you feel ready, that&#x2
131.
by s1t5 6y ago
> I guess it can lead to bad surprises. Surprises? If you build something and explicitly let anyone use it, you shouldn't be surprised when someone does in fact use it.
132.
by s1t5 6y ago
That kind of confusion has nothing to do with the terms though, it just requires a deeper understanding of the process and the statistics around it.
133.
Ask HN: How do you decide on what projects to work on at the intermediate level?
3 points by s1t5 6y ago|0 comments
134.
by s1t5 6y ago
Specificity and sensitivity feel really poorly chosen and difficult to remember. You could swap their definitions and the names of the terms would make just as much sense. I much prefer true positive rate and true negative rate.
135.
by s1t5 6y ago
> It's more like a paradigm changer, like the print, the engine, electricity, communication and computing. Either we really disagree about deep learning, or you vastly underestimate the influence of the other technologies that you&#
136.
by s1t5 6y ago
> That makes me think that the only ethical way of enforcing a temporary lockdown would be to also transition to a temporary marxist system. Collect 100% of profits and incomes and evently distribute it across the population whilst the e
137.
by s1t5 6y ago
None of them. Vitamin D is fat soluble and it's possible build up toxic levels of it. At the same time many studies suggest that humans need much higher doses than typically recommended, and some studies suggest that spaced hyperdosing
138.
by s1t5 6y ago
Isn't each of the hundreds of cryptocurrencies an alternative? Many of them have wildly varying implementations - different designs, levels of security, speed of transaction, transaction number limits, no blockchain at all, proof-of-st
139.
by s1t5 6y ago
Nice talk, you're a really good speaker. Thanks.
140.
by s1t5 6y ago
Only midlly related - for the past couple of years (since the crash of early 2018) following the crypto markets has been incredibly boring because of how highly correlated all the coins have been. Just have a look at coinmarketcap - the lin
141.
by s1t5 6y ago
I still don't get why multiple dispatch was chosen over having seperate functions with typed arguments. It just seems to add complexity with limited benefit.
142.
by s1t5 6y ago
Looking at the examples, I still struggle with code that imports multiple libraries at the top and then uses naked function names without telling me where those functions come from.
143.
by s1t5 6y ago
Even if we assume that saving that much is possible, the math only works if you're up for living a lifestyle costing 33% of your current earnings for the rest of your life.
144.
by s1t5 6y ago
> some terrible jokes I'm immediately shorting everything you own.
145.
by s1t5 6y ago
Good quiet office at a walking distance from my home, with dedicated desk spaces, dual monitors and respectful colleagues? - 100% office and 0% WFH. My own pre-virus situation of having to take public tranport in a big city, "hot-deski
146.
by s1t5 6y ago
Each ticket is a specific course/book/tutorial so it's easy to know if I've completed it or not.
147.
by s1t5 6y ago
I have a learning Trello board with categories "Maybe", "To Do", "In Progress" and "Done". When I come across something new it goes in the maybe category. I also make sure to only work on something if
148.
by s1t5 6y ago
> For home loans, the collateral is the house, which is strange because the house doesn't truly belong to the borrower until the loan is paid off. Your thinking is circular - the house doesn't truly belong to the borrower exact
149.
by s1t5 6y ago
Keep in mind that research in sports science is biased towards equipment like bikes because they make the study so much easier - safer to use, ensure consistent form, easier to get approval for the study and there's no need to coach th
150.
by s1t5 6y ago
In addition to all other arguments - it's up 500 points in the last month after dropping over twice as much the month before. The virus is still a net negative for the market.

More ›